Adaptive level generator in Super Mario Bros

This project aims at a novel procedural content generation method to generate high quality yet adaptable content which adapts games to match challenges that a game player can cope with. In this project, constructive primitives (CPs), quality yet controllable game segments in SMB, are learned via active learning. By means of CPs, we propose a dynamic difficulty adjustment algorithm that controls a CP-based level generator to adjust the content difficulty rapidly based on players' real-time game playing performance. For details on the adaptive level generator, please refer to our publication:

Shi P. and Chen K. (2018): Learning constructive primitives for real-time dynamic difficulty adjustment in Super Mario Bros. IEEE Transactions on Games 10(2): 155-169.

For demonstration, the executable jar file and tutorial of adaptive level generator are available below.

After downloading and unzipping the adaptive_generator.zip file, one should read tutorial documents in the tutorial folder. If there are any issues with the executable file and tutorial provided, please contact me (shipa@cs.manchester.ac.uk).